1. Field of the Invention
The present invention relates to a method of allocating identifiers for use in a distributed system using a LAN based on a broadcasting communication method such as a typical ETHERNET local area network and an apparatus for the same.
2. Description of the Related Art
A distributed system using a local area network (LAN) is typical of distributed or parallel processing systems having a plurality of computers. In such a distributed system, the physical distances between computers (sites) constituting the system are limited to a relatively short distance. However, since the system of this type not only provides a communication line with a sufficient capacity but also allows use of most existing techniques, systems can be easily constructed. For this reason, the distributed system using a LAN have been developed in various places and put to practical use. In a distributed system, an identifier such as a site number, i.e., a unique number assigned to each site, is generally used to identify each site. Under this site identifying method, it is required that a new identifier is allocated to a new site when the new site enters the system.
Various conventional methods of allocating identifiers have been proposed. Those conventional identifier allocation methods are based on the assumption that the individual sites are physically connected to each other with n(n--1)/2 number of lines where n is the number of sites and, for example, that site i can communicate only with site j by specifying a communication port of the site j through one of the lines.
However, since, in the distributed system using a LAN based on a broadcasting communication method such as ETHERNET, each site broadcasts the same information to the other sites, the above assumption will not hold. Therefore, conventional identifier allocation methods cannot be applied to LAN-based distributed systems. This is because, in such distributed systems, new sites to which no identifiers have been assigned enter into the system almost at the same time, and it is impossible to inform each new site separately of the identifier assigned to that new site through the communication line. That is, in conventional identifier allocation methods, only one site to which no identifier has been assigned can exist at the same time in the distributed system. For this reason, in the system using the LAN based on the broadcasting communication method, for example, newly entering sites have been assigned identifiers beforehand and the users activate the desired sites using the assigned identifiers. This conventional method requires much time to allocate the identifiers and also raises some problems such as issuing duplicate identifiers and vulnerability to unauthorized entering.